Abstract: Data processing in spark discharge atomic emission spectrometry （SD-AES） included the establishment of calibration curves, instrument standardization, and type standardization. Among these, the establishment of calibration curves and type standardization were classified as the calibration processes for instrument measurement data, while instrument standardization was the correction process for instrument drift （including the influences of personnel operations and environmental factors）. The consistent correction of data drift in routine analysis was also regarded as a correction process. In the SD-AES analysis system, linear interpolation and least squares method were mainly adopted for calibration and correction. The former was applicable to single-point and two-point correction, while the latter was suitable for correction with three or more points. The two-point clamping method was effective in solving the calibration problem of lacking suitable type standardization samples. For single-point type standardization, coefficient calibration was adopted in the low-content range, and addition calibration was used in the linear range. The broken line method realized the processing of wide-range type standardization. The segmented calibration technology based on the combination of broken line method and least squares method improved the precision of wide-range calibration curves and enhanced the universality of calibration curves for materials with the same matrix. In terms of instrument drift correction, the multi-point standardization based on weighted least squares method had stronger data restoration capability than the traditional two-point standardization, which could significantly improve the precision of SD-AES analysis.